1. Field of the Invention
Once the user has configured which call paths to monitor, VOIP automatically creates all the required IP SLA operations, including turning on the responder on the destination side. When these operations disappear, either due to their periodic expiration or due to a router reset, so the service recreates them as necessary.
2. Description of the Related Art
In voice over internet protocol (VOIP), telephone communications are sampled, digitized, and transmitted over a data packet transmission network over nodes in a distributed network to provide telephone communications between the nodes. VOIP is becoming increasingly popular due to improvements in performance and excess data transmission bandwidth that allows VOIP communications at relatively low costs. A complete description of VOIP communications are beyond the scope of the present application, but more information regarding VOIP communications can be founding in either ITU H.323, version 6 or IETF Session Initiation Protocol, the subject matters of which are hereby incorporated by reference in full.
In order to provide acceptable VOIP communications, the connections between the various nodes in the distributed networks must be configured. For example, the VOIP communications require sufficiently high QoS to allow data transmission without significant delay or data loss to avoid audio delay or audio loss. Typically, the communications between the VOIP network nodes are defined by service level agreements that define the default communication protocols between the nodes as needed, for example, to maintain the desired QoS levels. The need for high QoS communications needs to be balanced against the network overhead of providing the high QoS communications. Therefore, the service agreements ideally define a high QoS between nodes connection used primarily for VOIP data transfers but a low QoS between nodes connection used primarily for other, non VOIP data transfers.
Thus, these connections between the various nodes are each manually configured. Even a small VOIP network may include several local area networks (LANs), each having numerous interconnected nodes. Consequently, configuring the numerous nodes can be a laborious process.
The ability to pinpoint faults is important in VOIP, and daily support and troubleshooting are typically two of the most difficult parts about running a VOIP system. For example, it is difficult to anticipate the impact of losing that particular router or switch, with the VOIP traffic being rerouted in response to this change. Conventional VOIP monitoring tools monitor a network to pinpoint where problems come from and can model what will happen if the network topology changes. Most of the tools monitor jitter, packet loss, throughput, volume issues, delay, and other quality of service issues from within the network and/or call center applications. Nevertheless, the monitoring of the VOIP transactions over a network remains a tedious process that entails significant costs and computational overhead.
Conventional solutions measure VOIP quality in a “reactive” fashion by tracking end user phone calls from a span port on a switch. However, a local switch port is a poor location to gather network-wide VOIP quality metrics because VOIP quality and performance are determined through the performance of the connection from end-to-end, and monitoring the performance at the end node does not provide adequate information on the performance of the network. Also, monitoring actual call data raises privacy concerns due to the monitoring of actual calls.
Internet Protocol Service Level Agreements (IP SLAs™) enabled by Cisco IOS® software enable VOIP system monitoring by creating and monitoring synthetic voice data traffic. Likewise, other competing products provide similar functionalities. In particular, the various VOIP network components, such as routers and nodes may create synthetic traffic and the transmission of this synthetic traffic may be followed to gauge system performance.
In the CISCO IOS® software that resides on various network routers and nodes, IP SLA is an included feature that allows administrators to analyze IP service levels for IP applications and services, including VOIP. For more information on IP SLA, please refer to the IP SLA user manual at http://www.cisco.com/application/pdf/en/us/guest/products/ps6350/c2001/ccmigration—09186a0080789b77.pdf. IP SLAs use active traffic-monitoring technology to monitor continuous traffic on the network to measure overhead network performance. Routers further provide IP SLA responders that give accuracy of measured data across a network by receiving the synthetic data and forming performance network statistics.
In particular, IP SLAs are often used to generate data which is needed by the Service Level Agreements to define the characteristics of a connection between two network components, such as two nodes. With IP SLAs or similar synthetic voice data distribution tools, routers and/or switches may perform periodic measurements to monitor the status of the VOIP network and to collect network performance statistics without intruding on actual voice calls. These statistics include MOS, jitter, network latency, packet loss and other important QoS metrics that provide detailed visibility into VOIP performance.
The use of synthetic traffic throughout the system avoids the above-described privacy concerns while still providing reasonably accurate system performance measurements. Nevertheless, the use of synthetic traffic also has shortcomings. The data produced by the measurements of the systematic traffic may be voluminous and difficult to process. Also, the creation and transmission causes significant system overhead due bandwidth during the transmission of the synthetic traffic. Also, the processing of the synthetic traffic to produce the performance measurements burdens the processors in the routers and other associated network components.
Furthermore, programming of the VOIP nodes for IP SLA monitoring can be tedious. As described above, even a small VOIP may have numerous nodes. Also, a user is faced with the decision of using long term or permanent IP SLA that may consume excess network resources and produce large volumes of data, or using short term IP SLA that may expire during a desired monitoring period.